Toroidal Variable-Line-Space Gratings:The Good, the Bad and the Ugly

Edward West', Ken Kobayashi', Jonathan Cirtain', Allen Gary'-,John Davis' and Joseph Reader'

'Space Science Office, VP62, NASA Marshall Space Flight Center, Huntsville, AL 35812'The University of Alabama, Huntsville, CSPAR, Huntsville, AL 35899

"National Institute of Standards and Technology, Gaithersburg, MD 20899

ABSTRACT

Toroidal variable-line-space (VLS) gratings are an important factor in the design of an efficient VUV solar telescopethat will measure the CIV (155nm) and Mg1I (280nm) emissions lines in the Sun's transition region. In 1983 Kita andHarada described spherical VLS gratings but the technology to commercially fabricate these devices is a recentdevelopment, especially for toroidal surfaces. This paper will describe why this technology is important in thedevelopment of the Solar Ultraviolet Magnetograph Investigation (SUMI) sounding rocket program (the good), thedelays due to the conversion between the TVLS grating design and the optical fabrication (the bad), and finally theoptical testing, alignment and tolerancing of the gratings (the ugly).

Keywords: UV polarization, Solar Magnetograph, Sun, Sounding Rocket, Toroidal gratings, TVLS gratings

1. INTRODUCTION

The Solar Ultraviolet Magnetograph Investigation, SUMI, has been reported in several papers since this program beganin 2000$ • 9.10 . The emphasis of this paper is to describe SUMI's Toroidal Variable-Line-Space (TVLS) gratings. Thesegratings help SUMI meet its scientific goals which require both high spectral resolution and high optical efficiency formagnetic field measurements in the vacuum ultraviolet wavelength band of the solar spectrum (the good).Unfortunately, the technology readiness level of these gratings has made their implementation difficult, especially for asounding rocket payload (the bad). Therefore, this paper emphasizes the problems and solutions that were developed touse these gratings in SUMI (the ugly).

Section 2 contains a short review of the scientific goals of SUMI and why this mission is important in the understandingof the 3D structure of the magnetic field on the Sun. The flight hardware that makes up the SUMI payload is describedin Section 3 with emphasis on those components that affect the TVLS gratings. Section 4 emphasizes the alignment,testing and optical modeling that were developed to optimize the performance of these gratings.

2. SCIENCE GOALS

This section summarizes the Importance 1,2 of direct magnetic field measurements at various heights in the solaratmosphere and how the SUMI sounding rocket program plans to verify that transition region magnetic fieldmeasurements are possible.

2.1 Magnetic field measurements in the t ransition region

The problem of how energy, stored in the solar magnetic field, is released to heat the corona and drive the dynamicphenomena of the outer atmosphere, flares and coronal mass ejections, remains unsolved. Neither theoretical ornumerical models nor current observations are able to provide a conclusive story. We have learned that a substantialfraction of this energy propagates outward into the interplanetary medium in the form of EM radiation, energetic

https://ntrs.nasa.gov/search.jsp?R=20090034843 2020-06-03T14:20:16+00:00Z

Table 1: This table shows a comparison of candidate lines thatare being developed for current and fnhure space-basedphotospheric, chromospheric and transition region vectormagnetic field measurements.

Spectrallines

X(inn)

g X2xgx 104

Height(kan)

Remarks

Transition RegionCIV 1.55.0 1.2 2.9 2200 SUMIM II 280.0 1.2 9.4 2000 SUMI

Chromos hereCaII 854.2 1.1 80.2 1319NaI 589.5 1.3 45.1 ?00

Photos hereFeI 630.2 2.5 99.2 2.50 Hinode

particles, and the solar wind. These constitute spaceweather and have been shown to affect theimmediate geo-space environment and createhazards for terrestrial and space assets. In thefuture, space weather will be a major concern forastronauts working on the lunar surface andtraveling through interplanetary space- The solarinput is the starting point for any model thatdescribes and predicts space weather- Thereforeunderstanding the processes that contribute to boththe steady and transient release of magnetic energyinto the solar atmosphere is a crucial element ofthese studies.

To understand the fundamental processes that leadto the explosive release of magnetic energy in solarflares and coronal mass ejections (CMEs), it isnecessary to observe and infer the topology of the

magnetic field and homer it evolves prior to andduring the energy release process. Success has

b 1 d th h t f f th ii 11een ac ue^-e m e c arac errza ion o e uvector field in the photosphere, where (3, the ratio of the gas pressure to the magnetic pressure, is >I - At higher levels inthe atmosphere (Figure 1) 3 where R<1, the magnetic field; through the Lorentz force, controls the structure anddynamics of the solar atmosphere; and rapid changes in its structure can produce energetic events. However,observations of the magnetic field at these higher levels are difficult, placing a serious limitation on our understandingof the physical processes occurring there-

The missing element is the ability to understand and describe quantitatively the transition layer where the field is nolonger pressure-dominated as in the photosphere. The need to understand the energy release processes was identifiedtwo decades ago', and was reaffirmed as the major conclusion of a workshop on the definition of the scientific goals for"Beyond Solar-B" (Moore, Davis, and Hathaway, 2001)5 . To understand this process, full vector magnetic fieldmeasurements at several different heights in the solar 104atmosphere, from the photosphere to the upper > 1transition region where the field becomes nearly 3 ------- Jforce free, are required- The vector field in the 10

photosphere and low chromosphere has beenf

-'"'measured with ground- based instruments. Space- 2

10 = Corona 100,0instruments.based instr unents such as the Michelson Doppler 2 M II 20,000KInterferometer (MDI) on the Solar & Heliospherie 101 _ < 1 CaII 10,000KObservatory (SoHO) have demonstrated the .1importance of continuous, distortion free,

0observations of the longitudinal magnetic field from 10 = - --space to in our understanding of the Sun's magnetic Chromosphere------•--•• --

- ^field. The Solar Optical Telescope (SOT) on the 1 1 _PhotosphereHinode and the Helioseismic and Magnetic Imager = FeI 5,000K > j(HMI) on the Solar Dynamics Observatory (SDO) _2spacecraft will extend our space based observations 10 -- -- "" '4' 3 2 1 0 t 2

1010 10 1(10 10 10to include photospheric vector magnetic fields.Beta /B2)Although these high resolution observations will (16-,tnKT

greatly advance our knowledge of the behavior of the Figure 1. Ratio of gas pressure to magnetic pressure ((3) as amagnetic field in the photosphere, they will shed funrction of height, for an assumed range of photospheric

little light on how the magnetic field higher in the magnetic field strength of 100 to 2500 G (Gary, 2001)3.

solar atmosphere transitions to a nearly force-free

state. This requires measurement of the field using spectral lines that are formed at temperatures above 20,000 °K. Thetwo spectral lines we have chosen for SUMI are 280mn in MgII and 155nm in CIV. These lines occur in the farultraviolet and are inaccessible from the ground. Together with the lines FeI, NaI and CaII, they form a quintet ofmagnetically sensitive lines chosen for a future mission called the Magnetic Transition Region Probe (MTRAP) 6 . Thedemonstration that the MgII and CIV lines can meet the requirements for the force-free region is a prerequisite for thismission and is one of the objectives of the Solar Ultraviolet Magnetograph Investigation (SUMI).

2.2 Goals for SUMI sounding rocket program

Measuring CIV has always been the driving force in the development of SUMI. This line is formed in the relatively thintransition region which simplifies its interpretation. However, the magnetic sensitivity, V a where k is the wavelengthand g is the Lande g factor, is low compared to lines in the visible and infrared. Table 1 compares the SUMI lines withlines that are being used in photospheric and low chromospheric magnetographs. Certainly the results from SUMI willimpact future missions as scientist try to understand the 3-D structure of the Sun's magnetic field.

3. DESCRIPTION OF SOUNDING ROCKET PAYLOAD

The Solar Ultraviolet Magnetograph Investigation (SUMI) began as a set of development pro grams to improve theefficiency of polarization measurements in the ultraviolet. This section describes the sounding rocket payload withemphasis on those components that affect the Toroidal Variable-Line-Space (TVLS) gratings.

As demonstrated by Table 1, polarization measurements in the ultraviolet require high spectral resolution due to thesmall magnetic sensitivity (k'g). Higher spectral resolution is required to resolve the Zeeman splitting. Therefore toreduce the length of the telescope while increasing the spectrograph (with improved resolution), a cold mirror Ritchey-Chretien telescope was chosen over a traditional Gregorian telescope (§3.1). To improve the photon and polarizationefficiency, a MgF, double Wollaston polarizer was selected as the analyzer so that simultaneous measurements oforthogonal polarizations could be made (§3.2). Section 3.3 will describe the Toroidal Variable-Line-Space (TVLS)grating technology that is used to achieve the high spectral resolution while reducing the number of optical elements.The last section (§3.4) describes SUMI's cameras.

3.1 Telescope optics

A solar telescope design must solve the thermal problems associated with direct solar viewing. The simplest solution isa Gregorian telescope with a field stop between the primary and secondary mirrors. While this reduces the thermal loadon the secondary by rejecting the unwanted light, the disadvantages are its limited field of view, a longer optical path,and a larger secondary for on-axis designs, which decreases the effective collecting area of the telescope. SUMI'sapproach for decreasing the thermal load on the secondary mirror is to use a Ritchey-Chretien telescope with specialdielectric coatings applied to the front surfaces of both the primary and secondary mirrors (Figure 2). These coatingsreflect only the narrow wavelength ranges around the CIV and MgII emission lines which results in a "cold mirror".i.e.,a "self-filtering" telescope. The rear surface of the primary mirror is figured and has an aluminum coating that reflectsthe unwanted radiation back through the telescope and into space. The advantages of this design are that the field ofview is not restricted (the whole Sun can be imaged) and, for a given instrument length, the smaller telescope sizeallows a larger spectrograph improving the wavelength resolution. The narrowband UV reflection coatings simplify thethermal environment, minimize infrared and visible light contamination of the spectral data, and act as a blocking filterfor the spectrograph. The measured reflectance of these coatings is shown in Figure 3. Although the MgII bandpassmay appear off-center, its reflectivity was tuned to minimize the 1" order contamination of 3100A light into the CIVmeasurement_ which is made in 2" a order.

3.2 Polarimeter and slit jaw camera

The focal plane optics following the telescope includes the polarimeter. spectrograph slit and slit jaw camera. Thepolarimmeter consists of a MgFz waveplate and a MgFz double Wollaston analyzer. Due to the low photon flux at CIV

DHR TOY DHRsecondary primary DHR mirror CIV CCD

256x1024Out-of-band Out-of-band 155n

A 310nm^ I+QCIV

C IV AL mirror 1 Q! 280nm

MgII DHRMg MgII A&B

/ CIV CCD 512x512

r/ (2nd order) 1+QMgII I - Q

(1st order)

188.2 cm ►^

IIP

Mg

1 Mgr

2440L/mm

s a ^a5 ox155 270 155 22mA 1"280 132 I+Q 280 53mA 1"

Waveplate Wollaston TVLSAnalyzer satin

SIDE -- —

Figure 2. Optical schematic showing the technologies developed in the SUMI sounding rocket program. The CIVwavelength is 155nm (1550A) and the MgII is 280nm (2800A). The side view shows the double Wollaston analyzerand the two TLVS gratings designated by the I+QNP and I-Q/HP beam paths. Two frame transfer 512x512cameras are used in the MgII measurement (NIgIIA/HP and MgIIB/VP). A single full fi-ame 256x1024 CCD is usedfor the CIV VP/HP measurements.

and the weak linear polarization levels, the waveplate was designed to measure circular polarization at CIV (270°retardance at 155010 and the full Stokes vector at MgII (132° retardance at 280010 9' 1 . During the flight SUMI willconcentrate on circular polarization measurements when the payload is above 200km (required height for CIVmeasurements) and will make the linear polarization measurements when the sounding rocket is below 200km.

The double Wollaston analyzer which is a polarizing beamsplitter is the most efficient UV polarizes in this wavelengthband'. Figure 4 shoves the orientation of the double Wollaston and the exitin g linearly polarized beams. The orientationof the linear polarization and the polarization reflectivity of the TVLS gratings were discussed in a previous paper12.The selected gratings maximize the reflectance of the vertical linear polarization (VP = +Q defined to be the E fieldaligned to the slit, spatial axis), and horizontal linear polarization (HP= —Q is perpendicular to slit, dispersion axis).

A dual-beam analyzer is very important in transition region magnetic field measurements since the CIV emission issubject to rapid changes in intensity' $ . Without simultaneous measurements of orthogonal polarizations, intensitycrosstalk can create false signals in SUMI's magnetic field measurements. Since both polarizations are observed, thetotal transmission is much higher than traditional reflective polarizers even at CIV (the cutoff wavelength of MgF, is

1.2

SUMI telescope mirrors

1.0

C 0.8

all

v

0.6

v 0.4

0.2

o.o

1300 1800 2300 2800 33000

Wavelength (A)

Figure 3. Dielectric High Reflectance (DHR) coating on SUMIprimary and secondary telescope mirrors. These coatingsact as a prefilters to isolate the CIV (1550A) and MgII(2800A) lines. The MgII reflectivity was tuned to minimizethe 1 51 order reflectance of 3100A into the 2nd order CIVmeasurement.

tAir—0--Vacuum

I

I }

I t

1 1 1 1 1 1 1

1111111Input

aoutput

chi--^ -------Figure 4. Orientation of the double Wollaston with the linear

polarizations exiting the analyzer. The upper image is thepolarization parallel to the slit (vertical polarization, VP);the lower image perpendicular to the slit (horizontalpolarization, HP).

Since the double Wollaston analyzer produces

two polarized images of the slit; two toroidalvariable-line-space (TVLS) gratings arerequired`-`, `- The sensitivity of the spectrographgratings to the polarization exiting the doubleWollaston was taken into consideration'. In ourinitial concept a quarterwave plate was placedbetween the double Wollaston prism and gratingsto convert the exiting linear polarization tocircular if the gratings were found to be verysensitive to linear polarization. Fortunately, a barealuminum TVLS grating can be used for the

vertical polarization (VP) and a MgF2 coated aluminum grating for the horizontal polarization (HP see Table 2).Section 4 discusses both the optical modeling and measurements made on these devices.

3.4 UV cameras

The SUMI cameras are based on the E2V back illuminated, bare silicon technology. This technology enhances the LNquantum efficiency but at the time SUMI was developed their design was limited to certain CCD stnactures since theywere not part of E2V's standard product line in 2005. For the slit-jaw camera, and the MgII cameras where high speedelectronic shuttering was required, a frame transfer 512x512 CCD array with 13 pin was selected. The CNcamera uses a full frame 1024x2.56 CCD array with 26 µm pixels. To simplify the electrical wiring between the cameras

1150A). Finally, the polarization resolution (<10-3) is higher and covers a larger wavelength rangethan traditional reflective analyzers which isimportant for SUMI's CIV (1550A) and MgII(2800A) measurements. The rotating waveplateallows cross-checks of the image subtractionscheme by flip ping the polarization signals fromthe Sun between the HP3/VP5 detectors. Forexample, with the waveplate fast axis (FA) at 4.50the VP5 detector would measure I+V for CIV andthe HP3 detector I-V; with the FA at 135° VP5would measure IN and HP3 I+V (see Figure 7 forVP5/HP3 CIV detector definition).

The spectrograph slit has a unique stricture. Theslit is cut into a 6mm thick diamond turnedaluminum substrate and is -- 3.7rnm long (3.5 arcminutes) with 0.5 x 0.5 mm (30 aresec) boxes oneach end. The width of the slit has been split intotwo sections, a 30µm (--1.6 arc seconds) and a60µm (--3.2 arc second). Due to the exploratorynature of the SLJMI measurements, the uncertaintyin the target at launch and the limited observingtime of the flight (6 minutes), this structure wasadopted to give SUMI a larger dynamic range andto ensure capturing enough photons to make

meaningful polarization measurements at CIV_

3.3 Spectrograph

Table 2. Polarization reflectivity measurements for the six gratings fabricated by Jobin Yvon for SUMI: 2007measurements. The gratings selected for the first flight are #3 for the HP (HP3) and #5 for the VP (VP5).

Grating number CoatingHorizontal Polarization HP Vertical Polarization VPCIti'

(2A order)MgII

(1"order)CIV

2Idorder)MgII

(1" order1. MIS* AL/MgF2 12% 39% 29°../a 52%2. CY110 AL/MgF2 11°0 33% 15°../a 38%3. CYI IN* I AL/MgF 2 (AL) 1 12% (16%) 41% (9%) 30% (17%) 54% (49%)4. CYl lP AL 16% 11% 18% 52%5. CY11R AL 12% 1 15% 20% 1 62%6. CYl lM AL 14% 1 16% 17% 1 61%

* Acton AL,/MgF2 coatings. Grating 3 « vas recoated by Acton after the alliMiMrm measurements showed a low HPreflectance at MgII (9%)

and the data system, a USB 2.0 interface is used. For ground-based testing and software development, a MgF 2 windowis mounted in front of all of the cameras. The measured quantum efficiency (QE) for these cameras is >50%".

The SUMI sounding rocket has two data systems to control the four USB cameras, the rotating waveplate and tomonitor the health and status of the SUMI payload. Each data system has two USB and two Ethernet ports on theprocessor board. Since the CIV measurements have the highest priority, the CIV data system is the master and drives thetinting and telemetry for the SUMI instrument and controls the MgII data system. All of the data from SUMI's camerasis stored within the payload on a 4GB solid state disk drive attached to each data system. The slit jaw camera isconnected to the second USB port on the CIV data system.

4. DESIGN, FABRICATION AND TESTING SUMI'S TVLS GRATINGS

Kita and Harada described the use of spherical VLS gratings in 1983 13 and Thomas extended this technology to toroidalsurfaces in 2003 L3 . Due to the wavelength dispersion of the double Wollaston analyzer, a toroidal surface was animportant development for SUMI's CIV and MgII measurements This section discuses the optical design andfabrication of the gratings (§4.1), and the optimization, modeling and testing (§4.2) of SUMI's TVLS gratings in thesounding rocket payload. The optical design and fabrication are grouped together because it took several iterations todevelop an optical design that would fit within a sounding rocket. After the TVLS gratings were fabricated, theirpolarization properties were measured and two of the six gratings were selected for SUMI's first flight. The original setwas grating 1 for the HP (HPI) and and grating 3 for VP (VP3). After a failure during vibration testing, a second set ofgratings was selected, HP3 and VP5. The reason for this selection was based on the MgII HP reflectivity measurementswhich were higher for the aluminum gratin gs with a MgF2 protective coating (Table 2). The MgF 2 overcoat reduces theloss in the CIV reflectivity due to the oxidation of the aluminum coating and; while there is some improvement in theCIV VP reflectivity, the bare aluminum gratings will be used for the VP measurements so that there is a backup MgF2/AL HP grating.

4.1. Design and Fabrication

The SUMI design was developed and fabricated at different times: (1.) Telescope 10 - 2000; ( 2.) Polarimeter9 - 2001 and(3) TVLS gratings14 - 2003. This paper will only cover the design and fabrication of the TVLS gratings.

Certainly the development of the TLVS technology was good news for SUMI. With TVLS gratings SUMI couldachieve its wavelength resolution while improving its optical efficiency with the elimination of relay optics in thespectrograph. The problem was that the TVLS technology was just being developed which made the transition betweenthe optical design and the fabrication an educational process. This "educational" development increased the cost andcreated delays in the SUMI sounding rocket pro grain.

C B Rocket Skins.------ ---------------------------'F

'G DL-------------------------------------------

---------- -------:'CIV

;TVLS;n

^ ;TVLS( I-Q) i ; (I+Q)HP3 VP5

EMgIIB -- MgIIA

Figure 5. The critical areas in SUMI mechanical alignment that required adjustments to the original optical design. SeeFigure 6 for a 3D model of the SUMI spectrograph.

4.1.1. Development of the SUIVII Optical Design

The TVLS gratings went through several optical design cycles. Since the fabrication of these gratings required a longlead time, the initial TVLS design was completed in July; 2003 so that the procurement could be awarded. As thefabrication began, the second phase required adjustments in the placement of the gratings and the fold mirrors in theSUMI payload. The first design change was the beam separation between the VP and HP optical paths . Thisadjustment required that the gratings be staggered so that there was sufficient separation between them that they wouldnot come into contact with each other during the l OG vibration tests (Side view of Figure 2). The second adjustment tothe optical design was positioning the fold mirrors. This optical design required an adjustment in the CIV HP3 and theMgII VP5 (Fin Figure 5) fold mirror positions which was achieved by a small difference in the tilt of the HP and VPgratings (y tilt in Figure 6). The placement of the CIV VP5 fold mirror within the payload (Q and the MgII HP3 foldmirror away from the beams exiting the double Wollaston (G) also had to be considered. Finally, the USB interface ofthe flight cameras was longer than expected. That extra length coupled with the cooling ports (B in Figure 5) for thecameras thermal electric coolers required a final adjustment to the optical design to keep the cameras away from therocket skins (B & E) while eliminating any vignetting of the beam between the double Wollaston and TVLS gratings(A-&- D). While some loss in resolution was expected as the optical design was adjusted, this resolution loss was smallcompared to the riling errors encountered during the fabrication of the TVLS gratings (Figure 7, 2005 data).

Primary

Fold + Mirror

Mirrors

' Y ti

DeventerRotatlon

MgIIA

MgIIB

Focus of camera

Blue/top grating: VP5Red/bottom grating: HP3

y iCIV PPV-

DW

Figure 6. Alignment adjustments for the TVLS gratings

4.1.2. Fabrication

To optimize the instrument's imaging performance while fitting it into a sounding rocket payload, the spectrograph wasdesigned to measure CIV in 2"d order and MgII in V order. The two TVLS gratings use the same ruling equation. Theruling only varies by --2% around the central groove density of 2.440 lines per um (Figure 7). These ruling variationscan be described by a third-order polynomial with horizontal distance (y) on the gratin g blank (in mm):

1 + a y + p Y2 +

Y y3 (Zemax equation for VLS ruling)

Po

where p(t) is the ruling as a function of distance from center of the grating, and po is the central ruling frequency(lines/µm). The ruling parameters for the SUMI gratings are: po=2.440, a=-1.2755e-4, /1=1.182e-9 and y =2.4729e-11.Both gratings have the same concave toroidal radii of curvature: 1500.000mm in the dispersion axis and 1601.792mmin the spatial direction.

In October 2004, the manufacturer shipped SUMI's gratings to the optics group at GSFC to verify their performance.While the eun atures of the toroidal

2.50 surfaces were correct, an error in thetranslation of the Zemax rulingequation required that the gratings be

2.48 re-ruled (Figure 7: original rulings,2005 red dash curve). This requiredthat the grating designer become veryfamiliar with the manufacturer'sholographic ruling equipment. Afterseveral iterations adjusting the rulingequations of the Zemax optical designand the ruling equipment, the TVLSgratings were re-ruled and finallydelivered to MSFC August 2006(Figure 7: 2007 black solid curve).During this redesign, an analysisstudy was done on the new Zemaxruling parameters which showed thata was the critical parameter inachieving SUMPS wavelengthresolution. Since a is the linearcoefficient in the ruling equation, theassumption was made that, if themanufacturer stayed within their errorbudget, a small decenter and tilt of thegrating could compensate for thaterror.

4.2. Testing and Modeling the SUMI spectrograph

This section describes the development of a special VUV test facility that would interface to the SUMI sounding rocketand provide the spectral lamps needed to align the TVLS gratings (54.2.1)and the optical model to help in the alignmentof the CIV and MgII lines onto the spectrograph cameras (54.2.2).

2.46

Ew

2.44

2.42

2.40+ N ^

2007 spot sive

—60 —40 —20 0 20 40 60Position along gracing (ruling axis)

Figure 7. 2007 VLS ruling (black line, Zemax parameters: a=-1.2755e-4, ,8=1.182e-9 and y =2.4729e-11.) and the 2005 VLS (red dashed line, a=-1.2714e-4,8.8364e-8 and y =1.555e-11).

SUMI soundin rocket VUV collimator

_ Polari met er,&

SpectrographctionSectj.

Monochromator

. ',edge Imo'`=

Figure 8. SUMI sounding rocket attached to the VUV collimator (gold telescope) with the Acton VN1504. The VM504 wasreplaced with a Carbon lamp for the CIV alignment and a PtNe lamp for the MgII tests.

4.2.1. VUV collimator — spectral alignment

Since each optical element represents a loss in photons, especially at CIV; the goal for SUMI was to isolate the CIV andMgII lines using as few optics as possible. This approach led to the development of the dielectric high-reflectivity(DHR) coatings on the telescope mirrors (§3.1, Figure 3) which act as a dual-line prefilter. While SUMI is efficient inisolating our LJV lines, traditional methods to align the instrument cannot be used. Therefore a custom vacuumultraviolet (VUV collimator) facility had to be developed that was within the budget constraints of our sounding rocketprogram.

Although the V[JV collimator cannot achieve SUMI's resolutions, it can be aligned with traditional instruments. Thiscollimator is an all reflecting system with aluminum./MgF 2 mirrors that are optimized for 120nm. After aligning theVUV collimator, it can be used to provide the vacuum UV wavelengths needed to check the alignment of the SUMItelescope and spectrograph.

Our original concept used an Acton VM504 monochromator for the spectral calibration of SUMI but SUMI's higherwavelength resolution (2440 lines/mm TVLS gratings compared to the 1200 line/mm Acton grating), V themonochromator's f0.lnm uncertainty in its wavelength drive and the uncertainty in the wavelength position of theActon deuterium lamp's emission lines led to the decision to use a carbon spark lamp for the CIV alignment (Section4.2.2) and the platinum neon (PtNe) lamp for the MgII (Section 4.2.3).

Platinum/Neon Hollow Cathode source CIV: 1548.2 and 1550.7

10`-

10`-

10-

1544 15461548 1550 1552 1554 1556Figure 9. Emission lines of a PtNe source" with the SUMI wavelength range (green box) and CIV emission lines (blue

and red) superimposed. The green box is set by the size of the E2V CCD. The 256x26 Vin pixels represents thewavelength axis of the camera and the 1024x261un pixels are the spatial. /slit axis which captures both polarizationsfrom the double Wollaston (see Figure 10) .Emission lines of a PtNe source with the SUMI field of view (greenbox) and CIV emission lines (blue and red) superimposed.

VP5

HP3

Z 1550.8 Z

t Z 1548.2 - Z

Figure 10. SUMI display screen showing the overlay of the Zemax spectra (Labeled Z) on the SUMI spectrograph cameras. Theimage on the top are inverted and is the same as the lower image. The CIV camera sees both the horizontal (HP3) andvertical (VP5) gratings (green and light purple) at the same time.

4.2.2. CIV alignment

The initial alignment of the gratings began with the MgII cameras and the PtNe source but the large number of lines ledto some uncertainty in identifying the emission lines in the MgII wavelength band (Figure 11). While the CIValignment requires a vacuum (<I 0' torr), the carbon spark lamp actually provided the best wavelength calibration sinceit produced the CIV emission lines with only a few lines near that CIV wavelength band. With special wrenches, the tiltof the gratings were adjusted in a vacuum so that the CIV lines were centered onto the CIV camera.

To help in the alignment, a Zemax model was developed with the primary wavelengths near the CIV (and MgII)wavelength band and the spectrograph's custom slit. The weight parameter in the wavelength data was used to simulatethe intensity differences of the various lines. A dummy line with a weight of 0.001 was created in the wavelength 1position that was the midpoint between the CIV (and MgII) emission lines. With this setup, a Geometric ImageAnalysis (GIA) file could be created to show all of the emission lines. The field of view of the Zemax model was set upto be twice that of the spectrograph cameras so that nearby emission lines could be seen and compared to the imagesfrom the cameras.

Figure 10 shows an overlay of the Zemax model with the images from the CIV camera. In normal operation the CIVspectra are stored into the green memory plane of the display while the Zemax model is stored in the red and blueplanes. Additional line plots (not shown) are used to align the CIV spectra onto to Zemax model. These locations areindicated by the lines added to the Zemax GIA image.

4.2.3. MgII alignment

With the ali gnment of the VP51HP3 gratings near the correct position, the MgII alignment, which can be done in air,was accomplished by developing a Zemax model of the strongest PtNe lines near the MgII lines. Figure 11 shows thePtNe source in the MgII wavelength range. The lines that can be easily resolved in the MgII cameras are >10'. TheMgII lines (red and blue lines in Figure 11) are overlaid onto the PtNe spectrum with the green solid box indicating the

.3(0.3)

2794.2 (0.5)2793.3 (0.1)

—^ 2792.0 (1.0)

MgII: 2796.4 & 2803.5 (2799.9)

atinum/Neon Hollow Cathode source MgII: 2796.35 and 03.53A

1

>1 10,1^' t

10'tL

1011

2792 2794 2796 2798 2800 2802 2804

Figure 11. Emission lines of a PtNe source 16 with the SUMI wavelength range (green box) and MgII emission lines (blueand red) superimposed. The green box represents the cor rect alignment and the dashed red box the alignment shownin Figure 12. The size of the green box is set by the 512 x 13 Inn pixels of the E2V camera.

wavelength field of view of the MgII cameras. The red dashed box was added to indicate the wavelengths that are seenin Figure 12. Figure 12 shows the Zemax model with its alignment grid and the PtNe lines observed by the MgIIA(HP3) camera. When correctly aligned the right overlay plot will show "white" slit images (black on the right image).Since the Zemax model also tries to match the intensity of the PtNe lines (using the Weight parameter), the real-timetest images will also be a shade of gray.

5. SUNIlVLARY

TVLS gratings offer the designer the option to maximize the optical efficiency and wavelength resolution with a singleoptical component. This is especially important when making magnetic field measurements in the UV where everyoptical element represents a loss in the optical efficiency of the system. Unfortunately, there was a gap between theoptical design tools that describe the variable-line rulin gs and the fabrication technology. This uncertainty must beunderstood and evaluated before manufacturing these gratings (3 year delay in our schedule). Althou gh TVLS gratings

Figure 12. The Zemax spectra (red/yellow) onto SUMI's MgIIA image. The Zemax model has twice the field of view of theMgII camera. The image on the left is the negative of right image (normal mode) to allow identification of the weaker PtNelines. The PtNe lines are between the two images and the MgII lines are listed in the top left image with the midpointwavelength in parenthesis.

are still custom devices, they are important for high resolution VUV spectral measurements and, as more of thesedevices are produced, the gap between fabrication and design should be reduced. While the packaging of the TVLSgratings within a sounding rocket payload was challenging and resulted in some loss in resolution, the budget, schedule,optical analysis tools and test equipment were the primary problems associated with aligning these gratings. While theCIV alignment went more smoothly than expected, the MgII slit images forced us to improve our optical models and re-examine our alignment procedures and tolerances to explain our MgII images especially in our vertical polarizationtests.

6.REFERENCES

[1] H. Uitenbroek, "Multilevel Radiative Transfer with Partial Frequency Redistribution," Astrophysical Joiti-nal. 557,389-398 (2001).

[2] H. Peter, "On the nature of the transition region from the chromosphere to the corona of the Sun," A&A, 374,1108-1120 (2001).

[3] G. A. Gary, "Plasma Beta above a Solar Active Region: Rethinking the Paradigm," Solar Phys., 203, 71 (2001).[4] R. Rosner, B. C. Low, and T. E. Holzer, "Physical processes in the solar corona," Physics of the sun, D. Reidel

Publishing Co, 135-180 (1986).[5] Moore, Davis, and Hathaway, "High-Resolution Solar Magnetography from Space: Beyond Solar-B,"

http://solarscience.msfc.nasa.gov/Beyond Solar-B.shtml, (2001).[6] J. Davis, E. West, R. Moore, G. Gary K. Kobayashi, J. Oberright, D. Evans, H. Wood, J. L. R. Saba and D.

Alexander, "MTRAP: The Magnetic Transition Region Probe," SPIE: Solar Physics and Space WeatherInstrumentation. 5901, 1 -12 (2005).

[7] Gurman, J. B., "The MG II H line in sunspot umbrae," Solar Phys., 90, 13 -15 (1984).[8] West, E. A., J. G. Porter, J. M. Davis, G. A. Gary, D. M. Rabin, R. J. Thomas, and J. M. Davila, "Oven-iew of the

Solar Ultraviolet Magnetograph Investigation," SPIE: Instrumentation for UV/EUV Astronomy and SolarMissions, 4139, 350-361 (2000).

[9] West, E. A., J. G. Porter, J. M. Davis, G. A. Gary, and M. Adams, "Development of a polarimeter for magneticfield measurements in the ultraviolet," SPIE: Polarization Analysis: Measurement and Remote Sensing IV, 4481,pp. 109-117 (2001).

[10] West, E. A., J. G. Porter, J. M. Davis, G. A. Gary, M. Adams, S. Smith and J. F. Hraba, "Optical characteristics ofthe Marshall Space Flight Center Solar Ultraviolet Magnetograph," SPIE: UVrEUV and Visible SpaceInstrumentation for Astronmy and Solar Physics, 4498, 101-110 (2001).

[11] K. Kobayashi, E. A. West and M. Noble, "Polarization measurements in the Vacuum Ultraviolet," SPIE:Polarization Science and Remote Sensing Il, 5888, 1 -12 (2005).

[12] K. Kobayashi, E. A. West, J. M. Davis and G. A. Gary, "Polarization measurements on SUMI's TVLS gratings",SPIE: Polarization Science and Remote Sensine III, 6682, 1 -12 (2007).

[13] T. Kita and T. Harada, "Use of Aberration-Corrected Concave Gratings in Optical Demultiplexers," AppliedOptics, 22, 819-825 (1983).

[14] R. J. Thomas, "Toroidal varied line-space (TVLS) gratings," SPIE: Innovative Telescopes and Instrumentation forSolar Astorphvsics. 4853.. 411-418 (2003).

[15] L. Poletto and R. J. Thomas, "Stigmatic Spectrometers for Extended Sources: Design with Toroidal Varied Line-Space (TVLS) Gratings," Applied Optics, 43, 2029-2038 (2004).

[16] NIST, "Platinum Neon Data," http://physics.nist.gov/PhvsRefData:'platinum/contents.html , 1-2 (2008)[17] E. A. West, J. G. Porter, J. M. Davis, G. A. Gary, M. W. Noble, M. Lewis and Roger J. Thomas, "The Marshall

Space Flight Center Solar Ultraviolet Magnetograph," SPIE: UV and Gamma-Rav Space Telescope Systems,5488, 801-812 (2004).